A N o v e l Type of Flavonoids: Flavonol Esters from Fern Exudates
Eckhard Wollenweber
Institut für Botanik der Technischen Hochschule Darmstadt
Jean Favre-Bonvin and M. Jay
Departement de Biologie Vegetale de l ’Universite Claude Bernard, Lyon I
Z. Naturforsch. 33 c, 831 — 835 (1978) ; received August 30, 1978
N o th o la e n a , Farinose Exudate, Flavonol Esters
Farinose exudates on fronds of N o th o la e n a species are known to consist of flavonoid aglycones.
These are predominantly methylated flavones and flavonols, sometimes chalcones and dihydrochalcones. Flavonols with unusual properties, excreted as m ajor farina components by some species,
have been found to be derivatives of methylated flavonols, esterified in position 8 with butyric or
acetic acids, respectively. These are a novel type of flavonoids. They occur as twin-pairs, some with
very great differences in proportions. Their presence in a group of species within the genus has
chemotaxonomic implications. T w o of the flavonols resulting from hydrolysis are new compounds.
Quite recently several reports appeared dealing
with the chemical composition of the farinose deposit
on the under surface of fronds of ferns belonging to
the genera Cheilanthes, Notholaena and Pityrogramma (fam. Polypodiaceae, subfam. Gymnogrammoideae). In Pityrogramma chalcones and dihydrochalcones predominate [1 ], whereas in the
two other genera they are found rather rarely [2 ].
Cheilanthes and Notholaena produce mainly more
or less lipophilic methyl ethers of flavones and
flavonols [3 ]. It has been mentioned previously [3]
that some samples of N. ajftnis (Mett.) Moore and
N . californica D. C. Eaton had been found to show
“ completely unknown, but most probably flavonoid
compounds” . Thanks to the support of some
pteridologists supplying material for these investiga­
tions in sufficient amount, it has been possible now
to isolate at least the major components. From the
farina of N. affinis we identified novel 2,-substituted
flavonols with high degree of methylation of ring A
[4, 5]. Most interesting, however, are the esterified
flavonols encountered in several species of Notho­
laena. Here we report on elucidation of their struc­
tures.
Materials and Methods
Fern material was collected in the following
places (collector’s names and voucher numbers in
brackets) : N. affinis (L. D. Gomez 4725) in Costa
Requests for reprints should be sent to Doz. Dr. Eckhard
W ollen w eber, Institut für Botanik der T H , Schnittspahnstr. 3,
D-6100 Darmstadt.
Rica, N . aschenborniana Kl. (D. J. Pinkava & T.
Reeves R4315 and R4316) and N.neglecta Maxon
(D. J. Pinkava & T. Reeves R4310B and R4320)
in Mexico, N. californica (T. Reeves R 5353) in
California, N . galapagensis Weath. & Svenson (A.
& H. Adsersens No. 1972) on the Galapagos Islands.
The powdery flavonoid layer can be dissolved by
rinsing the fronds with acetone and benzene at room
temperature, sometimes with addition of methanol.
The compounds investigated are in part major com­
ponents of the farina and were precipitated from
concentrated extracts. They were purified by repeat­
ed crystallisation from methanol. NA-2 and NA-3
were isolated from the residue of N. affinis (after
crystallisation of NA-1) by column chromatography
and preparative thin layer chromatography. For CC
we used Polyamid SC-6 (Macherey & N agel);
elution was with benzene and increasing amounts of
methylethylketone and methanol. Preparative TLC:
Separation of NG-1 and NG-2 as well as of NA-1
and NA-2 was performed by TLC on Polyamid
D C -ll with solvent A, or on silica gel with solvent
benzene/dioxan/MeOH 90 : 12 : 3. Also NAS-1 and
NAS-2 were separated and NA-3 was isolated by
preparative TLC on silica gel. Solvents for poly­
amide: A ) benzene/petrol100_ 140°/methylethylketone/
methanol 30 : 60 : 5 : 5 or B) 60 : 26 : 7 : 7 or C)
benzene/methylethylketone/methanol
60 : 26 : 14.
Thin layer chromatograms were evaluated in long­
wave UV-light, before and after spraying with
“ Naturstoffreagenz A ” (/?-aminoethyl ester of di­
phenyl boric acid). Acid saponification: A solution
of the substance in a small amount of boiling glacial
acetic acid was treated with some drops of con-
E. W ollen w e ber et al. • A Novel Type of Flavonoids: Flavonol Esters from Fern Exudates
832
centrated hydrochlorid acid. The reaction product
precipitated while the solution was cooling down.
Total demethylation [6] : 2 mg of the substance were
mixed with 20 mg pyridine-HBr and heated on an oil
bath at 250 °C for 5 minutes. After cooling the
reaction mixture was treated with water and acetone
and extracted with ether. Mass spectra were recorded
using a Varian M AT 311 A at the institute of organic
chemistry of the Technische Hochschule Darmstadt,
PMR-spectra were recorded using a Varian X L 100
and peak-matching was performed using an
AEI MS 902 at the Universite Claude Bernard
Lyon I.
Results
All the compounds dealed with here are charac­
terised by the fact that they rather easily precipitate
from concentrated solutions, although they are dif­
ferent in polarity (see Table I ) . They crystallise as
light yellow needles. On polyamide layers they ap­
pear orange-yellow in UV-light, e. g. just like
flavonols kaempferol and quercetine. None of them
is identical, however, with one of the many reference
substances at our disposal. After spraying with
“ Naturstoffreagenz A ” they do not turn greenishyellow, but rather fluoresce yellow. Also their
molecular weights are not in accordance with the
structures of “ normal” , i. e. OH- and OCH3-substituted flavonoids. Hence they must be compounds
with unusual substituents. The reaction products
obtained on treatment with hydrochloric acid are
more polar than the original products. This leads to
the assumption that they are esters.
NG
This “ substance” , observed at first in the farina
of N. californica, after fruitless efforts to obtain
bulk material of this species, finally was isolated
from the exudate of N. galapagensis. On TLC in
non-polar solvents it can be seen that in reality it is
a “ twin-pair” , the components of which can be
separated only by preparative TLC. The substance
causing the lower spot (NG-2) could be isolated
from the farina of N.neglecta as a major constit­
uent.
Two peaks at mle 370 and 342, observed in the
mass spectrum of NG, appear as M+ in the mass
T a ble I. R f -v alues and UV-spectra of the flavanol-esters and the products of saponification ( “ sap.” ).
Substance
7?/X100
Solvent
B
C
A
62
83
N G -2
51
78
—
N G - sap.
12
31
78
8
25
(8 8 )
5
—
19
85
N A S - sap.
4
39
N A -1
48
81
N A S -2
N A -2
N A - sap.
N A -3
in M e O H
+ N aO A c
+ N aO A c
+ H 3B O 3
+ A IC I 3
+ A IC I 3
+ HC1
+ N aO H
366, 266
374, 266
368, 268
426, 334,
274
426, 334,
274
418, 268
390, (3 0 2 ),
280
375, 271
375, 272
450, 350,
(3 1 5 ), 268,
246
450, 346,
(3 1 3 ), 283,
246
370, 268
375, 324,
270
384, 268
376, 268
438, 356,
(3 1 0 ), 270
440, 354,
268
426, 344,
268, 250
388, 332,
277
383, 276
380, 336,
276
452, 365,
316, 283,
268
452, 365,
320, 282,
266
376, 284,
250
373, (3 2 0 ),
268
(4 0 8 ), 384,
(3 1 8 ), 268
372, (3 18),
268
435, 355,
(3 0 8 ), 268
435, 353,
(3 0 7 ), 268
418, 264
etwas
instabil
_
N G -1
N A S -1
U V-spectra (2max [n m ])
43
77
8
26
77
388. 330,
(3 0 8 ), 276
385, (3 2 0 ),
276
380, (3 42),
277
448, 369,
285, 270
450, 362
366, 248
instabil
13
41
(91)
3 7 8 ,(3 3 0 ),
(3 0 4 ), (2 7 0 ),
257
(4 2 0 ), 392,
258
378, (2 90),
(2 7 0 ), 256
438, 362,
266
436, 360,
267
422, 268
E. W o llen w e ber et al. • A Novel Type of Flavonoids: Flavonol Esters from Fern Exudates
spectra of NG-1 and NG-2, respectively; m/e 300
as the base-peak and lower fragments are identical
in both spectra. Here we provide data only on the
product of saponification, m/e 300 corresponds with
a “ normal” flavone or flavonol with 3 hydroxy
groups and 1 methoxy group.
The UV-spectra of both substances are completely
identical as are the reactions to the classical re­
agents [7 ]. The reaction with A1C13 and A1C13/HC1
confirm what had been assumed by the colour
development on polyamide: they are flavonols with
free OH-groups at C-3 and C-5. Furthermore the
spectrum in MeOH points to a flavonol of the
galangine type [7 ], i.e. with an unsubstituted 13ring. Position 7 is blocked; presumably the 0methyl group is located here. The PMR-spectrum
indicates an unsubstituted B-ring, too, and this is
indicated also by the presence of m/e 105 ( “ Pic
C” [8 ]) in the mass spectrum. Besides we observe
a signal for H-6, but not for H-8. The PMR-spectrum
furthermore shows the presence of a terminal
— CO — CH3 group.
Fragmentation of M-42
(NG-2) is in accordance with loss of an acetyl unit;
M-70 (NG-1) corresponds to loss of a butyryl unit.
Indeed on acid hydrolysis of NG-1 the smell of
butyric acid can be detected. Determination of the
molecular weights by high resolution mass spectro­
scopy leads to 370.1061 for NG-1 (calculated for
C20H i8O7, 370.1052) and 342.0739 for NG-2
(calculated for C18H 140 7 , 342.0739) and 300.0634
for the hydrolysed compound (calculated for
Ci6H i20 6, 300.0645).
The aglycone crystallises from the reaction
mixture of acid hydrolysis in fine yellow needles,
m.p. 246 °C. On polyamide TLC it forms a weak
brownish spot at much lower Rf than the original
compounds, which turns reddish brown on spraying
with “ Naturstoffreagenz A ” . Again the PMR spec­
trum shows an unsubstituted B-ring, a proton at
C-6 and one methyl group. The UV-spectrum indi­
cates a free hydroxy group at C-8 (bathochromic
shift of / maxH 24 nm, compared to N G ) .
MS m/e (rel.in t.): 300 (100), 285 (13), 271
(8 ), 269 (2 ), 257 (14), 183 (2 ), 167 (2 ), 155
(3 ), 153 (3 ), 139 (13), 111 (5 ), 105 (19), 77
(18). With regard to all the data available this prod­
uct must be the 7-methylether of 8-hydroxy-galangin.
Demethylation yields a product of still lower R f .
It appears as a brown spot on polyamide, which
becomes dark after spraying (bluish in daylight).
833
The solution after addition of ethylate turns blue —
a reaction which recalls the “ gossypetin-test” [9 ],
a hint for the 3,5,7,8-OH-substitution pattern.
Considering all the analytical results, NG-1 is the
butyric acid ester and NG-2 is the acetic acid ester
of 8-hydroxy-galangin-7-methyl ether.
NAS
Here we also have a mixture of two components,
isolated from the farina of N. aschenborniana. These
compounds are similar in behaviour to NG-1 and
NG-2, but have much lower Rf values. Again the
UV-spectra for both substances are identical. The
arguments for structure elucidation are about the
same as for NG and hence may be reported rather
briefly. Differences are in higher molecular weights
(386 and 358, base-peak at m/e 316) and “ Pic C”
[8] at m/e 121, behaviour of the UV-spectrum with
alkali (comp. Table I) and in the presence in PMRspectrum of the A A BB' spin system typical of psubstituted aromatic ring. From these observations
it can be concluded that the B-ring has an OH-group
at C-4/. So we probably have the esterified 7-methyl
ether of 8-hydroxy-kaempferol. Peak-matching results
in a mass of 386.0999 for NAS-1 (calculated for
Cooh i808, 386.1002) and 358.0695 for NAS-2
(calculated for C18H140 8, 357.0689), 316.0680
for the aglycone (calculated for C16H120 7,
316.0583).
Acid hydrolysis of NAS yields a yellow crystalline
compound, m. p. 285 °C (dec.). The maxima in
UV-spectrum differ by a few nms from those cited
in the literature [1 0 ]; there is no doubt, however,
that this product is identical with herbacetin-7methyl ether. MS m/e (rel.int.) : 3160 (100), 301
(20), 287 (6 ), 285 (2 ), 273 (11), 259 (3 ), 183
(4 ), 167 (3 ), 153 (3 ), 139 (10), 121 (17), 93
(5 ). Demethylation of the saponification product
yields a more polar compound. This appears as a
brownish spot on polyamide, becoming dark after
spraying (bluish-violet in daylight). Here also the
solution on addition of ethylate turns blue. Direct
comparison with an authentic sample of herbacetin
confirms the identity of both substances.
NAS-1 is the butyric acid ester, NAS-2 is the
acetic acid ester of herbacetin-7-methyl ether.
NA
The substance NA-1 has been isolated as a major
product from the farina of N. affinis. Elucidation
E. W ollen w eber et al. • A N ovel Type of Flavonoids: Flavonol Esters from F ern Exudates
834
T a ble II. P M R -sp ectra of some flavonol esters and of flavonols obtained by saponification
( “ sap.” ) . (100 M H z,
D M S O -D -6 ; <5ppm /TM S).
Flavonol esters
Flavonols
N G -2
N G - sap.
2.40 3 H 5 C O - C H 3
3.91 3 H
3.93 3 H s - O C H
6.54 1 H s H
6 .6 8
1
H
5
7.60 3 H m
H
—
3
5
-O C H
—
3
6
7.58 3 H m H - 3 ' 4 ' 5 '
6
H - 3 '4 '5 '
8.06 2 H dd H - 2 ' 6 '
(/ 8.5 and 2.5 H z)
8.23 2 H dd H - 2 ' 6 '
(/ 8.5 and 2.5 Hz)
A^4S-sap.
3.92 3 H s - O C H
6.56 1 H
5
H
—
3
6
6.95 2 H dd H - 3 ' 5 '
(/ 8.5 and 2.5 H z)
8.15 2 H dd H - 2 ' 6 '
(/ 8.5 and 2.5 H z)
N A -1
7V,4-sap.
1.02 3 H t - C H
(/ 7 H z)
3.88 3 H
3
1.75 2 H sext. - C H 2(J 7 Hz)
2.70 2 H t C O - C H 2(J 7 Hz)
3.85 3 H s - O C H
3
3.91 3 H
3
6 .6 6
5
-O C H
1H s H
—
-O C H
3
3.93 3 H s - O C H
3
5
6.59 1 H s H
— 6
7.16 2 H d H - 3 ' 5 '
(/ 9 Hz)
8.27 2 H d H - 2 ' 6 '
(/ 9 H z)
6
7.10 2 H d H - 3 ' 5 '
(/ 9 Hz)
8.03 2 H d H - 2 ' 6 '
(/ 9 H z)
small amount of the related substance NA-2. This
was not sufficient for a PMR-spectrum. All other
findings, however, above all in MS (M + 372, base
peak 330), allow its identification as the acetyl
compound corresponding to NA-1.
The product of saponification had not been
described in reference [11]. It crystallises in yellow
needles, m.p. 226 °C. Evaluation of the UV-spectrum
confirms once more the proposed structure as 7,4 dimethyl ether of herbacetin. MS m/e (rel.int.) : 330
(100), 315 (46), 301 (12), 300 (6 ), 299 (4 ), 287
(14), 153 (4 ), 135 (24). Certainty comes from
complete demethvlation, which again yields herbace­
tin.
When working up the remainder of N. aj finis
extracts a further yellow spot was observed on TLC.
A small amount of the product could be isolated by
preparative TLC on silica gel. The molecular weight
of 416 (base-peak at m/e 346 = M+-70) indicates
the presence of an additional OH-group, compared
with NA-1 /NA-2. The compound also is more polar.
Interpretation of the mass spectrum and of the UVspectrum and the PMR-spectrum shows that this
product NA-3 is the butyryl ester of gossypetin7,4 -dimethyl ether. In the mass spectrum a small
peak is observed at m/e 388, which indicates the
presence of the corresponding acetyl ester, NA-4.
Saponification and demethylation could not be
performed in this case because of lack of material.
N A -3
1.00 3 H t - C H
(J 7 Hz)
Discussion
3
1.74 2 H sext. — C H 2 —
(J 7 Hz)
2.72 2 H t C O - C H 2(/ 7 Hz)
3.85 3 H s - O C H
3
3.89 3 H s - O C H
3
6 .6 6
1H
5
H
—
The investigations presented here have shown that
certain Notholaena-species are able to exude not
only flavonoid aglycones but also esterified flavonoid
aglycones. Hitherto we found monoesters with
butyric acid and with acetic acid, respectively. All
the compounds of this kind are flavonols with a
6
7.11 1 H d H — 5'
(J 8.5 Hz)
7.56 2 H m H - 2 ' 6 '
(/ 8.5 and 2.5 H z)
of its structure as 8-butyryl ester of herbacetin-T,^dimethyl ether has been reported recently [11 ]. In
Tables I and II the data of this compound are
included to allow direct comparison with the analo­
gous compounds. From the remaining of the extract
from the isolation of the initially mentioned polymethoxy-flavonols it was possible to isolate a very
bzw.-C — CH-,
II
0
NG : R2=R3=H
NAS:R2=H , R3:0H
NA-1/2:R2=H, R3:OCH3
NA-3/j,:R2=OH, R3=OCH3
E. W ollen w eber et al. • A N ovel Type of Flavonoids: Flavonol Esters from Fern Exudates
methoxy group at C-7, esterified at C-8. The pro­
ducts differ by substitution of ring B.
These very stable butyric acid esters are the first
representatives of a new type of flavonoids. Substitu­
tion with an isoprenoid side-chain occurs rather
frequently [1 2 ]; its formation is biogenetically
evident. The butyryl side-chain is present in phloroglucinols from the fern genus Dryopteris [13], but
there it is C-C-linked. Recently an O-prenylated dihydrochalcone has been reported from Lonchocarpus
neuroscapha [14] and a new isoflavone with 0-3methyl-2-enyl substitution has been isolated from
seeds of Milletia auriculata [15]. True esters with
butyric acid, however, have not been reported
heretofore.
Esters of acetic acid already found in nature are
3-acetate of pinobanksin from bud exudate of Populus species [16], and its 7-methyl ether from seeds
of Alpinia japonica [17 ], and 2-methyl-7-hydroxy8-acetyl isoflavone reported recently to occur in roots
of Glycyrrhiza glabra [18].
Among the flavonol aglyca found here as esters
the gosspetin-7,4'-dimethyI ether has been known
only from seeds of Xanthoxylum acanthopodium
[19] .The herbacetin-7-methyl ether “ pollenitin” has
been reported from pollen of Camellia sinenesis [10]
and from leaves of Athraphaxis pyrifolia [20] (pre­
sumably glycosides in every case). The 7 ,4 -dimethyl ether of herbacetin and the 7-methyl ether of
8-hydroxy-galangin to our knowledge are new
natural flavonols.
The following survey shows the distribution and
relative amounts of flavonol esters present in farina
of the ferns analysed here.
These statements are valid only for the collections
cited here with herbarium numbers, i. e. for certain
E. W ollen w eber, Z. Pflanzenphysiol. 78, 344 (1976).
E. W ollen w eber, Z. Naturforsch. 32 c, 1013 (1977).
E. W ollen w eber, Z. Pflanzenphys. 85, 71 (1977).
M . Jay, E. W ollen w eber, and J. Favre-Bonvin, Phyto­
chemistry, in press.
[5 ] M . Jay, J. Favre-Bonvin, and E. W ollenw eber, Canad.
J. Chem., in press.
[ 6 ] G. H ow ard and T . J. M abry, Phytochemistry 9, 2413
(1 970).
[7 ] M . Jay, J.-F. Gonnet, E. W ollenweber, and B. Voirin,
Phytochemistry 14, 1605 (1975).
[ 8 ] H. Audier, Bull. Soc. Chim. France 1966, 2892.
[9 ] J. B. Harborne. Phytochemistry 8 , 177 (1969).
[10] Y . Sakamoto, A g r. Biol. Chem. 33, 818 (1969).
[1 1 ] E. W ollen w eber, J. Favre-Bonvin, and P. Lebrcton,
Phytochemistry 17, 1684 (1978).
[1 ]
[2 ]
[3 ]
[4 ]
835
NAS
NA
2
1
1 2
N o th o la e n a affinis
•
•
N o th n la en a aschenborniana
■
■
■
■
■
■
•
•
NG
1
N o th o la e n a californica
N o th o la e n a galapagensis
N o th o la e n a n eglecta
■
■
•
2
3
4
•
•
populations and must not be generalised. We have
good reason to assume the existence of chemotypes
with, in part, totally different flavonoid patterns
[3 ]. Butyryl esters and acetyl esters obviously
always occur jointly, although in very variable
amounts.
The study of farinose species of the genus Notholaena indicates that there are additional species
which are able to produce the novel flavonol esters
described here (Wollenweber, unpubl. results). After
further investigation of more specimens their distri­
bution will be reported in detail. The assumption is
that there exists a certain taxonomic grouping within
the genus which is characterised by this biosynthetic
capability.
E. W. wishes to express his sincere thanks to Dr.
H. Adsersen (Copenhagen, Danmark), L. D. Gomez
P. (San Jose, Costa Rica) and T. Reeves (Tempe,
Ariz.) for collecting and supplying the fern mate­
rials. He also is obliged to the Galapagos National
Park Service for permission to collect N. galapagensis. Thanks are due to Prof. Dr. H. Wagner
(Munich, GFR) for a sample of herbacetin and to
M. Petiaud (Lyon, France) for running the PMR
spectra. Prof. Dr. N. Higinbotham’s (Darmstadt,
GFR) kind help with correction of the English
manuscript is greatly appreciated.
[12]
[13]
[1 4 ]
[1 5 ]
[16]
[17]
[18]
[19]
[20]
K. Venkataram an, The Flavonoids (J. B. Harborne,
T. J. M abry, and H. M abry, E d s .), p. 277. Chapman
and Hall, London 1975.
C. J. W iden , J. van Euw, and T. Reichstein, Helv.
Chim. Acta 53, 2176 (1 970).
A . Lupi, G. D elle Monache, F. D elle Monache, and
G. B. M arini-Bettolo, Farm aco Ed. Sei. 30, 449 (1975).
K. V. Subba R aju and G. Srim annayarana, Phytochem­
istry 17, 1065 (1978).
E. W ollenw eber, Z. Naturforsch. 28 c, 638 (1973).
J. G ripenberg and K . Silander, Chem. Ind. 1955, 443.
D. K. B h ardw aj. T. R. Seshadri, and R. Singh, Phyto­
chemistry 16, 402 (1 977).
J. B. Harborne, P. Lebreton, H. Com bier, T. J. M abry,
and Z. Hamman, Phytochemistry 10, 883 (1971).
T. K. Chum balov, M . M . M ukhm ahed’yarova, and V.
B. Omurkamzinova. Khim . Prir. Soedin. 1974, 793.